Claude can now help you build evaluations and hillclimb on them.
In this article, we share guidance on eval design & skills that Claude Code can use to improve your applications.
https://t.co/PgKFC2DWth
Introducing Claude Sonnet 5.5, the second model in the Claude 5.5 family.
It’s a clear upgrade over Sonnet 5, runs more than 30% faster, and costs up to 30% less for most work.
I'm finding that agent loops that work off of GH issues and PR's with tags etc. are the best "factory" experience.
Novel idea, using the tools like they were intended right?
@daniel__designs@RobertJBye Sorry, yes.
So, I had not created a new remote session in like the last week and there must have been some updates. I think the only thing that it is still missing right now is being able to add sessions to a group. Sorry for the trouble, this is MUCH better
@localhost_5173 I think it makes more sense to build and validate with quicker languages and then you'll have the mold for the agents to refactor vs having to build, validate, wait yourself.
@gr36680@rushicrypto I bought a model Y. Still have our suburban and access to a 3/4 ton duramax. Rarely tow these days though, so it wasn't a particular issue for me. I do hope other automakers take inspiration from Tesla
@RobertJBye I did a demo in store of one. Not a huge fan of Vision Pro. Too heavy and I get motion sickness a little too easy.
I struggle to see the uses past gaming. The field of view wasn’t wide enough to make it feel better than proper monitors imo. Other than being much more portable